This study examined the relationship between individual and community-level factors concerning MHC service uptake (ANC, SBA and PNC). The results revealed that the uptake of these three MHC services was low in this study area compared with the national target set for 2020 [22], and with other African nations’ MHC service indicators [8]. An estimated 63% of the pregnant women in the present study used ANC services, which was lower than the corresponding result of nearly 70% in rural areas reported in EDHS 2019 [9]. However, the finding is consistent with a pooled point prevalence of ANC utilisation in Ethiopia, which was 63.7% [23].
The proportion of women who had given birth at healthcare facilities was estimated at 46.5% in the present study, but all the births that occurred outside of healthcare facilities were assisted by family members/TBAs, who may not have the skills and resources to manage life-threatening complications that can arise during labour and the immediate postpartum period [24]. Furthermore, the high number of home births could be a contributor to poor maternal health in the district. Therefore, relevant initiatives should be implemented to promote the use of SBA to help reach the national target of 90% SBA nationwide [22]. PNC use (32.8%) reported in this study is slightly higher than the 28.8% that EDHS [9] reported for rural women, as well as percentages from other studies conducted in Ethiopia [25] and Zambia [26]. However, this finding is consistent with a point prevalence from a systematic review in Ethiopia (32%)[27] and in a study from Nigeria [28]. The difference could be due to the fact that we included all visits to the health facility after childbirth, including health checkups and immunisations, under the assumption that every visit to the health care facility will encourage women to seek care at the time of contact. However, overall use remains very low, which is likely due to the high number of home births in this rural district [29].
The multilevel analysis found that individual factors, as well as community-level ones, significantly influence utilisation of MHC services in the study area. At the individual level, frequency of ANC and PNC visits were found to be influenced positively by knowledge of obstetric danger signs. This finding is consistent with previous studies that found knowledge of danger signs during pregnancy had a statistically significant association with the utilisation of MHC services [27,30,31], possibly because increased awareness of potential risks during pregnancy and delivery may induce behavioural changes, such as health-seeking behaviour, which includes seeking professional care [8]. Thus, working to raise awareness of danger signs during pregnancy and postpartum, as well as educating women about MHC services’ importance, would enable women to take advantage of existing services in this rural district [32].
Moreover, the likelihood of ANC visits also was significantly higher for women with planned pregnancies, which is consistent with findings from studies conducted in other parts of Ethiopia [11,30,31,33]. One possible explanation could be that women with unplanned or mistimed pregnancies devote less attention to the pregnancy and the care required for it [23,34]. It also is agreed that unplanned or mistimed pregnancy as a determinant in infrequent use of ANC has received less attention than other individually related barriers [35,36]. This study suggested that efforts to promote family planning among rural women to minimise unintended pregnancy not only may increase MHC service uptake, but also may reduce unsafe abortions, which is one of the country’s leading causes of maternal mortality [35,37].
Compared with housewives, women who work outside the home (farmers/labourers) were more likely to use SBA and PNC services. This finding was consistent with other studies conducted in Ethiopia [18,29] that also found employment is likely to enhance women’s status, helping them develop greater confidence to make decisions about their health, such as accessing healthcare facilities. Moreover, employed individuals were more likely to overcome financial constraints, which are a typical obstacle to obtaining MHC services in rural Ethiopia. Also, working women have easier access to information, which helps close knowledge gaps and creates positive attitudes that encourage women to seek MHC services [11]. Thus, creating local opportunities from which women could benefit financially is important to improving access to existing services in the area [38].
The use of ANC was found to be a significant predictor of the use of SBA [11,13,14] and PNC [25,27,39–41]. ANC offers pregnant women the opportunity to learn about preparing for birth and the benefits of childbirth in a healthcare facility, which, in turn, can influence their decision to use SBA [31]. While we could not find any significant difference in BPCR practice concerning SBA use, having a personalised BPCR plan that will help women prepare for potential emergencies during the childbirth process is advisable. However, not all women who used ANC also used SBA. The women may be reluctant to seek assistance from health care institutions if they believe their pregnancies are normal, or may experience access difficulties, including labour that begins late at night, when the women couldn’t get transportation to healthcare facilities [11,13,14]. Mothers who had given birth at a healthcare facility and had used ANC services were more likely to get counselling about PNC and danger signs after childbirth, which may encourage them to consider using existing services [25,27,39–41].
Women living in the lowlands were more likely to use ANC services compared with those in the highlands, who were more likely to have poor access to health services, poor infrastructure and longer distances to travel for health care. Moreover, a sociocultural difference such as traditional belief systems, access to education and wealth status also might hinder ANC use among highlanders [42]. Similarly, women living in Midland areas were more likely to use PNC than highlanders. The poor socioeconomic status of women in the highlands, as well as the presence of more health services in the lowlands, could explain this result. However, this finding also points to the need for a contextual and localised intervention that will benefit all women in the area [12,43].
According to a systematic review from developing countries, distance and increased travel time to the nearest health care facilities have been found to be associated significantly with ANC and SBA use. Similarly, ANC and SBA uptake was lower in communities where the healthcare facility was difficult to reach [21,43,44]. Consistently, this study found that having easy access to a health facility made a considerable impact on ANC service use, as limited access to health facilities negatively affects health-seeking behaviour and may have resulted in low ANC use. Similar to most developing countries – in which most infrastructure is concentrated in urban areas and is scarce in rural areas, where most of the population lives – the disparity makes it difficult for women, particularly those living in rural areas, to access health care. While efforts aimed at closing the gap between rural and urban areas are encouraged, working on improving individual women’s access is suggested [21,44].
The study also indicated that women who lived in communities with health centres (BEmONC) within two hours in travel distance and who had easy access to transport were more likely to use SBA. Other studies also found similar results elsewhere in the same context [44,45] and in developed countries [46]. Women who must travel for more than two hours and who had difficulty accessing transport were less likely to give birth in a health centre. This result suggests that a physical accessibility problem exists for rural women who want to use existing healthcare facilities. Many women in rural areas must walk long distances on difficult terrain or be carried on a traditional stretcher to reach a health centre. This journey through the hills under difficult geographical and health circumstances leaves many women either opting to give birth at home or giving birth on the side of the road when they cannot reach a facility in time, likely increasing the risk of complications or death for both the mother and the unborn baby [47].
Making these services physically accessible through the establishment of maternity waiting homes (MWHs), which are facilities that house pregnant women during the final few days or weeks of their pregnancies, provides easier access to nearby healthcare facilities and is one of the common practices in developing countries, including Ethiopia [48,49]. However, even though MWHs are one of the mitigation strategies for addressing disparities, most MWHs suffer from a lack of quality care and were not integrated into the health care system in Ethiopia [49,50]. As a result, integrating MWHs into the health system while maintaining an explicit link to the community should make it easier for women living in rural and remote areas to access facilities [51].
Strengths and Limitations
This study’s sample was drawn from intervention and comparison areas for the evaluation of a community-based intervention to improve the use of MHC, which was implemented after a complete census of pregnant women in Gamo Zone’s Arba Minch Zuria district in Southern Ethiopia. Previous research has focussed on one or a few aspects of MHC services, but we evaluated the three key MHC indicators and tried to identify determinants at individual and cluster levels to provide a complete picture of maternal health, as well as valuable information for policymakers that can be used when planning context-specific interventions.
However, the findings should be interpreted in light of several limitations. Considering that the data were self-reported, particularly distance and travel time, this was prone to recall and social desirability bias, which could have influenced the data’s internal validity. Recall bias was reduced by focussing on the most recent birth during the past five years, and the interviews were conducted in private places to reduce social desirability bias and reassure participants on data confidentiality. Moreover, causal inferences are not possible with observational data examined in this study. Furthermore, the proxy used for distance to health facilities did not measure actual distance, but rather derived it from respondents and data collectors’ estimates. The extent to which this proxy variable truly reflects the distance to services is uncertain.